View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Northumbria Research Link Citation: Elsweiler, David and Harvey, Morgan (2015) Engaging and maintaining a sense of being informed: Understanding the tasks motivating twitter search. Journal of the Association for Information Science and Technology, 66 (2). pp. 264-281. ISSN 2330-1643 Published by: Wiley-Blackwell URL: http://dx.doi.org/10.1002/asi.23182 <http://dx.doi.org/10.1002/asi.23182> This version was downloaded from Northumbria Research Link: http://nrl.northumbria.ac.uk/22211/ Northumbria University has developed Northumbria Research Link (NRL) to enable users to access the University’s research output. Copyright © and moral rights for items on NRL are retained by the individual author(s) and/or other copyright owners. Single copies of full items can be reproduced, displayed or performed, and given to third parties in any format or medium for personal research or study, educational, or not-for-profit purposes without prior permission or charge, provided the authors, title and full bibliographic details are given, as well as a hyperlink and/or URL to the original metadata page. The content must not be changed in any way. Full items must not be sold commercially in any format or medium without formal permission of the copyright holder. The full policy is available online: http://nrl.northumbria.ac.uk/policies.html This document may differ from the final, published version of the research and has been made available online in accordance with publisher policies. To read and/or cite from the published version of the research, please visit the publisher’s website (a subscription may be required.) Engaging and Maintaing a Sense of Being Informed: Understanding the Tasks Motivating Twitter Search DAVID ELSWEILER1 and MORGAN HARVEY2 1Chair of Information Science Institute for Information, Media, Language and Culture University of Regensburg 31. Universitätsstrasse, 93053 Germany Telephone: +49941 943-1708 Telefax: +49941 943-1954 Email:[email protected] 2 Faculty of Informatics Università della Svizzera Italiana (USI) Via Giuseppe Buffi 13 CH-6904 Lugano, Switzerland Telephone: +41 58 666 43 00 Telefax: +41 58 666 45 36 Email: [email protected] Abstract Micro-blogging services such as Twitter represent constantly evolving, user-generated sources of information. Previous studies show that users search over such content regularly, but are often dissatisfied with current search facilities. We argue that an enhanced understanding of the motivations for search would aid the design of improved search systems, better reflecting what people actually need. Building on previous research, we present qualitative analyses of two sources of data regarding how and why people search Twitter. The first, a diary study (p=68), provides descriptions of Twitter information needs (n=117) and important meta-data from active study participants. The second data set was established by collecting first person descriptions of search behaviour (n=388) tweeted by twitter users themselves (p=381) and complements the first data set by providing similar descriptions from a more plentiful source. The results of our analyses reveal numerous characteristics of Twitter search that differentiate it from more commonly studied search domains, such as web search. The findings also shed light on some of the difficulties users encounter. By highlighting examples that go beyond those previously published, this article adds to our understanding of how and why people search such content. Based on these new insights, we conclude with a discussion of possible design implications for search systems that index micro-blogging content. 1 Introduction Twitter is a socially-focussed short messaging (“micro-blogging”) service that allows users to post and read short messages - known as “tweets” - of up to 140 characters in length. In these tweets users post about what they are currently reading, thinking and doing and often post URLs to web sites of interest to them (Java et al., 2007; McFedries, 2007). Twitter is incredibly popular and is increasingly embedded in everyday life. As of June, 2013 Twitter has 218 million monthly active users, who collectively send around 500m tweets a day (Rushe, 2013) and as of February 2012, some 18% of online american adults use Twitter and users of the service have diverse demographics (Brenner and Smith, 2013). In addition to being a platform for socially sharing thoughts and opinions, work has shown that Twitter also represents a valuable, user-driven source of information of unprecedented volume (Boyd et al., 2010). Tweets can provide “specific information, useful links, and insights from personal experiences” (Hurlock and Wilson, 2011). Previous work has shown that many tweets are questions directed to the user’s followers (people who sign up to receive tweets from that user) in the hope that they can provide an answer (Morris et al., 2010). Twitter also offers an oft-used search interface to publicly available tweets (Lin and Mishne, 2012) and major search engines have recently started to include appropriate tweets as separate verticals in their search results, highlighting the importance of the medium. Teevan et al. (2011) report 126,000 searches from 33,000 users over a period of just two weeks and Lin and Mishne (2012) reveal that up to two billion requests are made to the Twitter search API every day. Despite the frequency of their use, there is some evidence to suggest that users are dissatisfied and frustrated by Twitter’s search features in their current form (Ingram, 2011), a fact that Twitter themselves are aware of and are taking steps to address (Shin, 2013). Query log analyses have shown that Twitter searches have very different properties to other kinds of search e.g. web search (Teevan et al., 2011; Lin and Mishne, 2012). Twitter information needs are often highly temporal with high levels of query churn (Lin and Mishne, 2012), but, at the same time, repeated queries are more commonly re-issued than on the web (Teevan et al., 2011). Furthermore, the social networking aspect of Twitter influences search behaviour with links between users (@ links) and topics (hashtags) being vital parts of the experience not present in other systems. While we know what searches on Twitter look like, very little research has been undertaken to understand the information needs behind these searches. We argue that if researchers were to understand more about the motivations behind Twitter searches – what do people want to find, why and what problems do they face – it would lead to the design of improved retrieval models and interfaces and, therefore, an enhanced user experience. Furthermore, knowledge of the kinds of information people want to find through Twitter could be valuable when designing services that make use of Twitter data. We add to this understanding by describing the results of a diary study designed to learn about the diversity of information needs that motivate searching Twitter content and a second collection of data from Twitter to confirm and expand upon the patterns found. The findings reveal numerous characteristics of Twitter search that makes it different to more commonly studied search problems such as web search, as well as some of the difficulties that can be experienced while searching. We report several findings of note and discuss the implications these have for the design of search systems that exploit micro-blogging data. 2 Related Work We structure the related work for this article into two parts. In Sub-section 2.1, we review background literature relating to Twitter to demonstrate the value of a study such as the one described in this article. In Sub-section 2.2, we summarise work from Information Seeking on classifying search tasks. This relates to the results of our investigation and provides a basis from which discuss our findings. 2.1 Twitter Research The popularity of Twitter has made it a topic of research interest in many fields. Prior work has evaluated the way in which Twitter is used to share information in various contexts, e.g. during elections (Gaffney, 2010) and natural disasters (Vieweg et al., 2010) and for different purposes, e.g. to engage particular groups of people (Boyd et al., 2010). Twitter content has also been used to understand sentiment (Pak and Paroubek, 2010), predict future trends (Bollen et al., 2011) and replace tags as information sources for URLs (Harvey et al., 2012). The field of Information Retrieval (IR) has also contributed significantly to social media research. Such work has shown that people frequently search over Twitter content (Lin and Mishne, 2012), although the properties of tweets, such as their short length, method of creation and short lifespan, means they are searched differently from web pages. People often want real-time search results (Teevan et al., 2011) and the most popular queries often reflect celebrities and news stories, both of which change in importance and relevance with time (Teevan et al., 2011). Freshness is an important concept in social-media search (Mishne and de Rijke, 2006) and, as such, many social networks order results in reverse chronological order (Thelwall and Hasler, 2007). In 2011 the TREC micro-blog track was launched as a platform to experiment with retrieval models for such media. Reflecting the organisers’ beliefs regarding search tasks, topics are currently grouped into three categories: News Categories, Geographical Interest and Topic Target (entity sought-after) (Soboroff et al., 2012). While these tasks are perfectly plausible, little research has been done to investigate different usages of Twitter search i.e. to understand what people are really trying to achieve. To our knowledge, the only previous work in this direction is a survey of 54 Microsoft employees on their Twitter search habits (Teevan et al., 2011). The findings emphasise the temporal patterns in Twitter search with Memes, Twitter user names, and celebrity names all being popular Twitter queries.